بررسي عوامل مؤثر بر قيمت مسكن شهري با استفاده از شبكۀ عصبي مصنوعي موردشناسي: منطقۀ دو تبريز

Assessment of Effective Factors on Urban House Prices Using Artificial Neural Network؛ Case Study: District 2 of Tabriz

ﻣﺴﮑﻦ، در ﺑﺴﺘﮥ ﻣﺼﺮﻓﯽ ﺧﺎﻧﻮارﻫﺎ ﺳﻬﻢ ﭘﺎﯾﻪ اي دارد. در ﺣﻘﯿﻘﺖ، ﺑﺮاي اﻏﻠﺐ ﺧﺎﻧﻮارﻫﺎ ﺧﺮﯾﺪ ﻣﺴﮑﻦ ﻣﻬﻢ ﺗﺮﯾﻦ ﺗﺮاﮐﻨﺶ ﻣﺎﻟﯽ آن ﻫﺎ ﻣﺤﺴﻮب ﻣﯽ ﺷﻮد. ﻫﻤﭽﻨﯿﻦ، ﻣﺴﮑﻦ ﺳﻬﻢ ﻗﺎﺑﻞ ﺗﻮﺟﻬﯽ از ﻫﺰﯾﻨﻪ ﻫﺎي ﺧﺎﻧﻮار و در ﺑﺮﺧﯽ ﻣﻮارد ﺣﺘﯽ ﮐﻞ داراﯾﯽ ﺧﺎﻧﻮارﻫﺎ را ﺗﺸﮑﯿﻞ ﻣﯽ دﻫﺪ. ﺑﺎزار ﻣﺴﮑﻦ، ﻣﯽ ﺗﻮاﻧﺪ ﺗﺤﺖ ﺗﺄﺛﯿﺮ ﻣﺘﻐﯿﺮﻫﺎي ﮐﻼن اﻗﺘﺼﺎدي، ﺗﻔﺎوت ﻫﺎي ﻓﻀﺎﯾﯽ، وﯾﮋﮔﯽ ﻫﺎي ﺳﺎﺧﺘﺎري ﺟﺎﻣﻌﻪ و اﻣﮑﺎﻧﺎت رﻓﺎﻫﯽ ﻣﺤﯿﻂ ﻗﺮار ﮔﯿﺮد. ﺑﺪﯾﻦ ﺳﺎن ﮐﻪ ﻧﺎﻫﻤﮕﻦ ﺑﻮدن ﻣﺴﮑﻦ و ﭼﮕﻮﻧﮕﯽ رﺗﺒﻪ ﺑﻨﺪي وﯾﮋﮔﯽ ﻫﺎي ﻣﺨﺘﻠﻒ ﯾﮏ واﺣﺪ ﻣﺴﮑﻮﻧﯽ ﺗﻮﺳﻂ ﺧﺮﯾﺪاران ﺳﺒﺐ ﺷﺪه اﺳﺖ ﺗﺎ ﻗﯿﻤﺖ ﻣﺴﮑﻦ دﺳﺘﺨﻮش ﺗﻐﯿﯿﺮات و ﻧﻮﺳﺎﻧﺎت ﺷﻮد. ﭘﮋوﻫﺶ ﺣﺎﺿﺮ، ﺑﻪ دﻧﺒﺎل ﭘﺎﺳﺦ ﮔﻮﯾﯽ ﺑﻪ اﯾﻦ ﺳﺆال اﺳﺖ ﮐﻪ »ﭼﻪ ﻋﻮاﻣﻠﯽ ﺳﻬﻢ ﺑﯿﺸﺘﺮي در ﺗﻌﯿﯿﻦ ﻗﯿﻤﺖ ﻣﺴﮑﻦ در ﻣﻨﻄﻘﮥ دو ﺗﺒﺮﯾﺰ دارد؟«. ﭘﮋوﻫﺶ ﺣﺎﺿﺮ ﺑﻪ ﻟﺤﺎظ ﻫﺪف، ﮐﺎرﺑﺮدي و ﺑﻪ ﻟﺤﺎظ روش و ﻣﺎﻫﯿﺖ، ﻫﻤﺒﺴﺘﮕﯽ اﺳﺖ. از ﺷﺒﮑﮥ ﻋﺼﺒﯽ ﻣﺼﻨﻮﻋﯽ ﺑﺮاي ﺳﻨﺠﺶ ﻫﻤﺒﺴﺘﮕﯽ ﺑﯿﻦ ﻣﺘﻐﯿﺮﻫﺎي اﺳﺘﻔﺎده ﺷﺪه اﺳﺖ. اﻃﻼﻋﺎت ﻣﺮﺑﻮط ﺑﻪ واﺣﺪﻫﺎي ﻣﺴﮑﻮﻧﯽ از ﻃﺮﯾﻖ ﻣﺮاﺟﻌﮥ ﻣﺴﺘﻘﯿﻢ ﺑﻪ ﻣﺸﺎوران اﻣﻼك ﺟﻤﻊ آوري ﺷﺪه اﺳﺖ. ﺟﺎﻣﻌﮥ آﻣﺎري، واﺣﺪﻫﺎي ﻣﺴﮑﻮﻧﯽ ﻣﻨﻄﻘﮥ دو ﺗﺒﺮﯾﺰ ﺑﻪ ﺗﻌﺪاد 56107 ﻣﺴﮑﻦ اﺳﺖ ﮐﻪ ﺑﺎ اﺳﺘﻔﺎده از ﻓﺮﻣﻮل ﮐﻮﮐﺮان، 384 ﻧﻤﻮﻧﻪ ﺑﺮآورد ﺷﺪ و ﺑﺮاي ﺑﺮآورد ﻣﻄﻠﻮب 400 واﺣﺪ ﻣﺴﮑﻮﻧﯽ ﺑﻪ ﺻﻮرت ﺗﺼﺎدﻓﯽ ﺑﻪ ﻋﻨﻮان ﻧﻤﻮﻧﮥ ﭘﮋوﻫﺶ اﻧﺘﺨﺎب ﺷﺪه اﺳﺖ. ﯾﺎﻓﺘﻪ ﻫﺎي ﭘﮋوﻫﺶ ﻧﺸﺎن ﻣﯽ دﻫﺪ ﮐﻪ ﺳﻬﻢ ﻣﺘﻐﯿﺮ ﻫﺎي ﮐﺎﻟﺒﺪي در ﺗﻌﯿﯿﻦ ﻗﯿﻤﺖ واﺣﺪﻫﺎي ﻣﺴﮑﻮﻧﯽ 53/8 درﺻﺪ، ﺳﻬﻢ ﻣﺘﻐﯿﺮ ﻫﺎي دﺳﺘﺮﺳﯽ ﺑﺮاﺑﺮ ﺑﺎ 39/2 درﺻﺪ و ﺳﻬﻢ ﻣﺘﻐﯿﺮ ﻫﺎي ﻣﺤﯿﻄﯽ 7 درﺻﺪ اﺳﺖ. از ﺑﯿﻦ ﮐﻞ ﻣﺘﻐﯿﺮ ﻫﺎ، ﻣﺘﻐﯿﺮ ﻫﺎي ﻣﺴﺎﺣﺖ زﯾﺮﺑﻨﺎ ﺑﺎ 28/4 درﺻﺪ، ﻓﺎﺻﻠﻪ از ﻣﺮاﮐﺰ درﻣﺎﻧﯽ ﺑﺎ 6/4 درﺻﺪ، ﻓﺎﺻﻠﻪ از ﻣﺮاﮐﺰ ﺑﻬﺪاﺷﺘﯽ ﺑﺎ 5/1 درﺻﺪ و ﻧﻤﺎي ﺳﺎﺧﺘﻤﺎن ﺑﺎ 4/6 درﺻﺪ ﺑﯿﺸﺘﺮﯾﻦ ﺳﻬﻢ ﻣﺘﻐﯿﺮ ﻗﯿﻤﺖ ﻣﺴﮑﻦ را ﺑﻪ ﺧﻮد اﺧﺘﺼﺎص ﻣﯽ دﻫﻨﺪ. در اين پژوهش از نرم افزارهاي MATLAB 2013 و ArcMap 10.4 بهره گرفته شده است.

Housing is a fundamental component of the household consumption bundle. In fact, for most households, the purchase of a home is their single most important financial transaction. The housing market can be influenced by macro-economic variables, spatial differences, characteristics of community structure, and environmental amenities. Heterogeneity of house and how consumers rank different characteristics of a house led to price changes and fluctuations. So that, one house with similar physical attributes in different urban regions will show different prices. This research, looking for recognition effective factors on house Prices and estimating prices of housing units in the District two of Tabriz. This research based on applied and Correlational researches. The data were collected through survey and inquiry from real-estate agents. Statistical population is the houses in district two of Tabriz which is 56107. Cochran formula estimated 384 sample size and for desirable estimation 400 house were randomly selected as a sample of research. Artificial neural network (ANN) is employed in this paper to analyze housing values. In determining the house prices physical variables have 53.8 percent, distance variable has 39.2 percent and environmental variables has 7 percent. The findings of research indicate which floor area variable with 28/4 percent, distance from treatment centers variable with 4/4 percent, distance from health centers variable with 5.1 percent and building facades variable with 4.6 percent has the highest share of house prices. In this research, were used MATLAB 2013 and ArcMap 10.4.